RLCA Observations from the Crow's Nest
by Dagmar Cole
The crow's nest sits high above the deck and provides the best view for a lookout to spot approaching hazards. RLCA positions the IT strategist in the crow's nest with the ability to see the existing landscape as well as the distant horizon.
Companies are focused on "next quarter earnings" and a one-year (or less) return-on-investment. It's difficult to sell upper management on a technique that takes staff years to complete and more than one fiscal year to payoff. How does RLCA compare to other SISP methodologies — Business Systems Planning (BSP), Strategic Data Planning (SDP), and Strategic Management Planning (SMP)?
Timeliness. RLCA can be done relatively quickly. Performing analysis on the Human Resources resource consisted of a total of 5,280 hours or approximately 2.5 man years. This is an order-of-magnitude less than performing other SISP techniques.
Quality. Both business and IT users were able to comprehend the RLCA concepts and assess the value chain. Using desktop tools, not only basic reports but graphics were created.
Maintainability. Existing RLCA does not require any specialized tools and can be performed using a spreadsheet or a desktop database. Existing tools can be modified to support RLCA and thus does not require additional cost or training.
Reproducibility. As each reviewer examined the resource value chain, results were concurred by both IT and business users. Given that other SISP methodologies have such a long time to complete, with the business landscape constantly changing, it is highly unlikely that any other SISP methodology can be realistically reproduced.
Usability. BSP, SDP, and SMP each have a tremendous chasm between the resulting strategic plan and the implementation. RLCA has a tight coupling to the enterprise strategic plan. The resulting precedence architecture provides the map needed for prioritizing applications development and as well as highlighting those applications to be retired. RLCA additionally identifies gaps that most likely are undocumented 'stovepipe' applications.
As an application architect for a Fortune 50 organization, I have implemented RLCA in my organization and the results have been phenomenal. Using a tool no more sophisticated than a desktop database, I had my business analysts identify the resources, value chains, and precedence. Using this as a basis, applications were mapped to the value chains, then entities. This was done as part of the normal requirements-gathering process and not as separate projects so there was "no cost" to the business.
Immediately there were major benefits;
- Clusters of applications supporting the same resource functions were identified.
- Some functions had no identifiable IT application support.
- The ability to compare the resource value chain with the process flows identified tasks that may not be providing value to the organization.
- It provided a basis for project prioritization.
- It provided the ability to accurately bound and scope the project's level of effort, resulting in business case estimates going from +/- 200% to /- 25%.
Strategically, RCLA identified the 'low-hanging fruit'. Functions with missing supporting applications were reviewed first. In all cases there was a spreadsheet or manual process that was automated, and we were able to identify 'quick hit' improvements that were low cost and high value to the organization. These quick hitters were used to gain the business' trust.
The RLCA precedence architecture provides an enterprise-wide view of the business and associated IT applications. RLCA identifies candidate redundant applications by viewing clusters of applications supporting the functions. Are the applications supporting critical or low-priority functions? If low priority, why so many? For example, in Organization 'A', where the pilot was performed, why were there over forty applications supporting Recruitment? After discussions with the manager, it was discovered that new employees wanted tools they were familiar with so tools were bought; then subscriptions were purchased for ongoing maintenance and support, with little to no use. A decision was made to reduce the number to three, resulting in immediate savings to the organization without any associated costs.
Conversely, viewing clusters of applications around high-priority projects identified potential projects to integrate. Often we found that there was a single function that was not integrated, which required the entire application to remain operational. One small functional sliver like this resulted in ongoing inflated maintenance costs. Knowledge is power — armed with this information, developing integration roadmaps became extremely easy and aligning IT with the business strategy was a breeze.
The precedence architecture also proved to be an invaluable tool for estimating a development project. Rather than the 30,000-foot view of architectural diagrams from an IT perspective, the precedence architecture provided a realistic map to bound the actual scope of the project. For example, we knew that changing an application for this resource function would impact ten other areas.
The result? Over a three-year period of the business tracking variations between the 'raw' order of magnitude (OOM) estimates and the actual final project costs, the variations were in the range of +/- 25%.
There is a saying, "There is nothing new under the sun." In this case, RLCA has been around, largely languishing, for nearly twenty-five years. I say it's time to dust it off and for it to take its rightful place in the sun.
DISCLOSURE: Dagmar Cole has no relevant relationship with Ron Ross or Business Rule Solutions.